CBD Ad Rules: What You Can and Can't Say in 2026
A practical breakdown of CBD advertising restrictions across major ad platforms and how to work within them without tanking performance.
CBD ad rules boil down to this: most major ad platforms (Meta, TikTok, Google) either restrict or outright prohibit ads for CBD and other hemp-derived products, and the rules vary by platform, by country, and by whether the product contains THC. There is no single universal rulebook — the safest approach is to treat every platform's policy as its own gate, avoid health claims entirely, and build creative that can survive a manual review without relying on the word "CBD" doing the selling.
Why is CBD advertising restricted in the first place?
CBD sits in a regulatory gray zone in a lot of markets — it's derived from cannabis, even when it's non-psychoactive and legally sold. Ad platforms default to caution because enforcement bodies hold them accountable for what runs on their networks, and because CBD products frequently get marketed with unproven health claims (curing anxiety, chronic pain, sleep disorders) that regulators actively police. Platforms would rather block a category broadly than adjudicate every individual ad's legitimacy.
What typically gets a CBD ad rejected?
Rejections cluster around a few patterns: any implication the product treats, cures, or prevents a medical condition; imagery or copy that emphasizes intoxication or "getting high"; landing pages that make the same claims the ad avoids; and, in some cases, the mere presence of the word "CBD" or a visible product shot of a tincture or vape. The specifics shift often, so the practical move is to check the current policy for each platform you're running on before you brief creative — not to rely on last quarter's rules.
Can you still advertise CBD at all?
In many cases yes, with restrictions — some platforms allow topical CBD products (balms, lotions) with age-gating and no health claims, while ingestible products face tighter limits or outright bans depending on the region. Some brands route spend through platforms with more permissive category policies, or lean on channels outside paid social entirely (SEO, email, affiliate). Whatever the mix, the constraint is the same: no disease claims, no guaranteed results, no cure language, ever.
How do you build ads that convert within these limits?
Since the loudest claims are off the table, the ad has to win on hook and craft, not on a promise regulators would flag anyway. Focus on the founder story, the ritual (how and when people use it), texture and format, and honest before/after framing that stops short of medical claims. Because that leaves less room to lean on a single silver-bullet script, it pays to test a wider spread of creative — different hooks, different actors, different pacing — and see what a compliant, claim-free ad can still do on watch time and hook rate.
Where does Polaris fit into this?
This is exactly the situation Polaris is built for: you stay inside AI UGC ads made from your own product photos, so there's no stock-avatar library implying claims you didn't write, and the actor delivers only the script you give it. Add the Polaris MCP connector to Claude once, describe the compliant angle you want to test, and generate a batch of variations in parallel instead of hand-producing one safe-but-flat video and hoping it's the right one. Check hook strength with the hook generator, and once ads are live, track whether the more restrained creative is actually earning its budget with the ROAS calculator — see ROAS and hook rate for the underlying definitions. For the full workflow, see generating ads inside Claude via MCP.